Talking Points for RSO-III
1) Intro slide
2) Objectives (as listed)
3) Example of the various air masses we will be reviewing. Note the cloud streets in
sly flow in the warm sector…i.e. an unstable boundary layer where the clouds are
aligned with the surface flow. A closer look shows and enhanced region of streets in
central NE. We will be talking about this type feature later. Also note the wave
clouds in the relatively cool stable boundary layer where the clouds are perpendicular
to winds at the top of the inversion. In between these air masses are the boundary
where convection may initiate.
4) Conceptual diagram of low-level reflectivity with low-level streamlines. Outflow
boundary on the southwest side (cold front) corresponds to leading edge of cooler air
being left behind by the storm. Green area represents an intense surge of possibly
cooler outflow air due to the development of a rear flank downdraft (RFD). The RFD
is a result of a) a downward-directed pressure perturbation force associated with mid
lvl dry air entrainment interacting with the maturing mesocyclone in a supercell (and
to a lesser extent the anvil high), and b ) evaporation of precipitation into the dry,
subsiding air.
5) Schematic of visible satellite image, showing external cloud features that often
develop in association with a supercell. Overshooting top (OST), above anvil
cirrus that forms in the warm wake. Associated with the storm scale cold front is a
flanking line. On the cool side of the flanking line we may observe lines of
towering cumulus developing above a new or intensifying RFD. On the warm
side of the flanking line where the inflow is converging into the strongest updraft
you frequently see compact lines of towering cumulus feeding into the storm.
These cloud lines are called “feeder clouds” by Dennis et. al, and Amand in the
early 1970’s. The development of these lines are often an indicator of intense inflow.
Their presence generally means that storm intensification is taking place.
6) Examples of supercells that exhibit external storm features referred to above. Frame
1) Hesston, KS tornadic storm (13 March 1990). Note similarity of features to
diagram on previous slide, Frame 2) Wichita Falls, TX tornadic storm (10 April
1979). Note large overshooting top, outflow boundary to west of the storm, enhanced
towering Cu just east of flanking line, and Frame 3) Plainfield, IL tornadic storm
(28 August 1990) where CAPES were approaching 6000 j/kg on this day.
7) GOES RSO information slide.
8) Review of Why RSO is important…most important – during regular scan, images
take about 25 min to arrive on AWIPS from the time image scan begins. With
RSO…images arrive on AWIPS about 8 minutes after image start time. Much
better temporal continuity with RADAR and other data sets.
9) Case I; 5 Aug 2000 CWAs
10) IR channel 4 loop from 0331 – 1815 UTC
Loop shows two MCS’s - one in SD moving southeast. Second MCS in ND moving
northeast. ND MCS produces a southward moving outflow boundary can be seen on
both IR and vis (shown later). SD MCS is moving through Iowa and is probably
leaving a stable region in its wake. In eastern South Dakota we can see a subtle
outflow boundary moving southward, we will analyze the visible imagery later to
gain continuity of this feature.
11) Omaha, NE (OAX) sounding with hodograph at 1200 UTC. Shows atmosphere
unstable, keep in mind this is early August, dewpoints are high (70s). Low-level
shear weak. Mid- to upper level winds are sufficient for severe storms. Will the
boundary layer develop stronger and backed flow?
12) Eta analysis MSLP contours (yellow), 500 Height contours (green) /sfc wind barbs
(yellow)/500 mb wind barbs (cyan) at 12:00 UTC. Shows backed winds over the
area of interest but weak sfc flow at this time. Trough line extends southwest from
the low in southeast South Dakota.
13) Eta CAPE/MSLP (yellow) /500 height (green) sfc wind barbs (yellow)/500 mb wind
barbs (cyan) at 18:00 UTC. Compare with surface obs at 18:00 UTC. Note the
maximum instability in southeast Nebraska / southwest Iowa. Trough line is moving
eastward into our region of interest.
14) Loop of surface observations 13:00 – 18:00 UTC. Note how high the dewpoints are
in eastern Nebraska / western Iowa. The area south of the MCS boundaries in eastern
Nebraska / southwest Iowa heat up into the 80s. Sfc winds in eastern Nebraska veer
as trough moves through. Wind speeds are very light, thus the primary threat appears
to be hail rather than tornadoes.
15) Visible satellite loop 14:00 – 20:45 UTC. MCS exiting IA leaves behind stable air
and boundary. A second boundary (from the ND MCS) is moving south through
eastern SD. Note small cloud field moving into southwest Nebraska. This cloud field
will be analyzed in the following slide to determine what height these clouds exist.
16) Output from AWIPS Cloud Height Algorithm at 20:45 UTC 5 August, 2000 – The
clouds in question in southwest Nebraska are analyzed using the AWIPS cloud height
tool on 10.7 um imagery. The clouds are found to exist around 480 mb, these are
mid-level clouds likely associated with a shortwave trough.
17) Neligh, NE profiler located in northeast Nebraska. Confirms that wind speeds aloft
are increasing during the afternoon. This is supporting evidence that the mid-level
cloud field of interest is associated with a short wave.
18) GOES-11 Single Field of View (SFOV) sounder CAPE 12:16 –20:46 UTC. This
product utilizes ETA forecast sounding as the first guess (GFS for real-time
products). The parcel is chosen by selecting the highest theta-e from lowest three
height levels.
For this time period, destabilization takes place west to east from eastern
Nebraska into western Iowa. CAPE values above 4000 J/Kg over region of interest.
Note cloudy area in southeast Nebraska. Pink values around clouds (higher CAPE
values) may be contaminated due to the diffraction near the clouds. Drier air lags
behind the wind shift along the trough line.
19) GOES-11, visible 1-min interval satellite loop 19:15 – 22:52 UTC. Notice the stable
wave clouds in northeast IA. This airmass seems to be moving east (we also
confirmed this by looking at the increasing GOES sounder CAPE values moving
towards the east). Notice southward moving boundary from earlier MCS approaching
unstable air in northwest IA. A second southward moving boundary becomes evident
ahead of the other southward moving boundary. This may be a result of insolation, or
the shortwave trough arrival, or both. Both boundaries play a role in convective
initiation.
Note also that the nose of the apparent shortwave is moving into eastern NE.
As the upper clouds move into central NE we see stable wave clouds develop
underneath. This tells us two things; a) the airmass in central NE is stable, and b) the
shortwave has a lot of forcing. The arrival of these clouds in eastern NE/western IA
seems to be associated with the onset of storms along the more northern southward
moving boundary at around 21:30 UTC, then along the southernmost southward
moving boundary at 22:40 UTC.
20) 0.5 degree tilt composite radar reflectivity with NLDN GG lightning toggle overlay.
Loop runs from 20:54 – 00:42 UTC. First notice that the second storm actually forms
on the southernmost southward moving boundary that could only be seen later with
daytime heating on VIS. The first storm dies. A third storm forms in approximately
the same location where the original storm died. When we toggle on the lightning
data we can see that the second storm is a PSD storm so it is most likely severe.
Higher flash count just before 3” hail at 23:45 UTC. The later, northern storm has a
high flash count from the onset.
21) KOAX (Omaha) 0.5 degree tilt radial velocity with NLDN CG lightning overlaid at
00:42 UTC. No clearly defined mesocyclones in either storm. Lots of large hail, no
tornadoes
22) GOES-11, 1-min interval, visible loop 22:53 – 00:47 UTC. Early in the loop. we
note that there are inflow feeder clouds, but no horizontal rolls west of the flanking
line. Shortly thereafter we can see stratiform clouds pushing out ahead of the storm.
Note that these clouds are casting a long shadow. The height of these clouds will be
analyzed on the next slide, where they are found to be mid-level clouds. Notice the
double pulse in the OST as it relates to the double pulse in the advancing stratiform
clouds. Remember that this is not a classic supercell – little directional shear, no
tornadoes reported. Note OST associated with northeastern storm appears at 00:22
UTC. On radar isolated we see the new echo at 00:18 UTC.
23) Output from AWIPS Cloud Height Algorithm at 23:45 UTC 5 August, 2000 – The
height of the clouds below the anvil level cirrus observed rushing southeastward from
the storm are found to be around 580 mb. These are mid-level clouds, not low-level
outflow. These are most likely caused by a gravity wave associated with the storms’
updraft.
24) 0.5 degree tilt composite radar reflectivity with NLDN GG lightning toggle overlay.
Loop runs from 00:48 – 03:18 UTC. Good correlation between decreasing flash rate
(beginning around 02:00 UTC) with storm dissipation at about 02:45 UTC.
25) Case Conclusions.
26) Case II; 2 Jun 1998 Pittsburgh, PA (PBZ), State College, PA (CTP), and
Binghamton, NY (BGM) CWAs. This case focuses on the differences between lowlevel airmasses, how we can identify them using satellite, and how that information
can help in the Nowcast/WDM process.
27) Eta MSLP and surface winds. a) analysis – deep surface low northern Lake MI,
apparent front back in IL, b) 18:00 UTC low moves to southern Lake Huron, two
troughs – note the pre-frontal trough that extends southeast to North Carolina. c) By
00:00 UTC low moves to Ottawa, with cold front in west-central PA.
28) Infrared satellite loop 10:01 – 13:15 UTC. Shower activity and cloud cover early
in the morning over eastern two-thirds of PA and most of NY associated with
vigorous early am shortwave. Where will baroclinic boundary(s) setup due to cloud
cover or the rain cooled boundary layer? What other products could help? Let’s look
at some model output and satellite imagery.
29) Eta CAPE helicity and Bunker right-mover vectors. a) analysis time, b) 18:00
UTC increasing CAPE from west, 200 helicity, c) 00:00 UTC CAPE axis (around
2000) through west central PA, best helicity on east side of axis
30) Visible loop 13:31 – 17:45 UTC. What does the imagery show us about various
airmasses? Where is the most unstable air at 16:15 UTC? Where is the stable air?
See the Cu developing early over southwest PA…then Cu forming later in MD and
southeast PA. Are both areas characterized by streets? The area of streets in
southwest PA develop first, so if both areas are streets, then the earlier development
may have been due to earlier heating and more rapid destabilization. In southeast PA,
cumulus clouds do develop later, but streets remain capped as they move north behind
boundary (warm front or baroclinic zone setup from early morning clouds/showers).
These are not wave clouds, because wave clouds would be stationary. Also, can tell
by looking at soundings.
31) 18:00 UTC soundings from Sterling, VA (LWX) in green and Pittsburgh, PA
(PBZ) in cyan. Areas both east and west of the Appalachian Mountains are both
unstable. Low-level flow parallel to streets and airmass very moist. Neither
sounding has a strong cap, but the PBZ inversion is much weaker…therefore, makes
sense that Cu starts to go up first ovr wrn PA vs ern PA.
32) Terrain map of area of interest. Note the ridge lines and try to relate to low-level
flow. Low-level southwest flow has an upslope component on the west face of the
ridge. From west-to-east we see the Chestnut Ridge, the Laurel Highlands, and the
Allegheny Front Range.
33) Visible loop 18:45 – 23:15 UTC. Thunderstorms first develop in northeast OH. We
can actually see two lines of storms go up…with the first line forming along the prefrontal trough, moving into/through the Pittsburgh CWA. Second line of storms
forms on the front in OH at around 19:00 UTC. Notice that the forcing is so strong
that deep convection is even taking place over Lake Erie (at this time of year lakesurface temperatures are typically around 50 F). There continues to be a richlooking, persistent area of cloud streets in southwest PA. By running the loop faster,
we can see persistent features better (streets, clearing region downwind of
Chesapeake Bay, persistent clear area in northeastern PA extending into NY). The
pre-conditioned airmass represented by the streets is feeding into the storms.
Remember that the orientation and movement associated with the cloud streets
represents the integrated shear vector in the boundary layer.
Note boundary in eastern PA. Which storms would you be most concerned about
from satellite alone (radar a little later). Hard to see under anvil into forward flank
convergence area. Second line of storms is following right behind first. Note that
storms form and move through eastern PA, confirming that this area was capped, but
unstable. Why does the southernmost storm in southeast PA dissipate between 2130
– 22:00 UTC, while the storm further north does not?...All due to the/a boundary!!!
Also note storm ovr wrn NY around 21 UTC…particularly the storm scale features
the the crisp anvil and feeder clouds/towers.
34) KPBZ 0.5 degree reflectivity 21:04 – 01:21 UTC with surface data overlaid. Draw
conclusions between sat and rad data, plus possible terrain effects. Note: first F1
tornadic storm – leaves behind boundary. Note merger just before this tornado
forms. Surface still southwest and low 60s dewpoints. Second storm produces an F4
tornado at 01:01. Note cloud line develops on terrain feature. Last storm F2 in
southwest PA. Wind shift behind second line.
F1
21:55 Southwest PA, first storm
F2
22:45 Western NY
F2
23:00 Southwest PA, first storm (may have been Derecho)
F4
01:01 Southwest PA/western MD second storm
F2
01:34 West VA
F3
02:00 Northeast PA
35) 3.9 m, shortwave IR loop (Channel 2) 20:32 – 23:15 UTC. We will now look at
the 3.9 m imagery since it is getting late in the day. We can see the first storm in
southwest PA lays out an outflow boundary NW-SE oriented which would be parallel
to storm motion for the next storm that follows. The second storm would be worth
watching more carefully for this reason. An E-W boundary is evident in northeast
PA. See next item for a few storm reports. This channel is good for using near
sundown and into night hours. Remind that with GOES R, resolution will be almost
as good as current visible images.
36) Conclusions. Remind…to use satellite data/images to supplement other data in
evaluating pre-storm environment…then continue to use into the warning
process to continue to show the near storm environment and to augment
decisions made with the RADAR.
37) Case III; 28 Apr 2003 Pocatello, ID (PIH) CWA. This case shows how terrain
can affect the evolving severe thunderstorm situation, and how knowledge of the
local terrain can help focus the short term forecast and nowcast.
38) Water vapor imagery 09:45 – 18:30 UTC 28 April 2003. Closed low off northern
California coast. One shortwave is ejecting across eastern Idaho. Second shortwave
can be seen moving across central Nevada. Diffluent flow across Idaho. Note that
we should expect to see these well organized shortwaves on the Eta analysis (shown
later). Jet moving into/through NV/AZ/UT region.
39) 10.7 m, IR satellite loop 12:00 - 18:45 UTC. Zoomed in closer to our CWA, note
that this channel is looking closer to the surface allowing us to see terrain features as
they relate to the approaching shortwave. Early in the loop we can clearly see the
colder terrain associated with the ridge lines throughout eastern Nevada . As the loop
goes on colder clouds can be seen developing on these ridge lines, as well as ridges in
southeastern Idaho, and advecting off towards the northeast. Cloud top temperatures
indicate warming tops in Oregon and developing cold tops approaching our CWA.
40) Terrain in S. Idaho. Snake River valley - multiple ridge lines in SE Idaho - Sublette
Range in the Sawtooth National Forest. The Snake River winds approximately east-
to-west across southern ID. The Snake River valley is called by different names in
different locations – Upper Snake in southeast ID, Magic Valley in south-central ID,
and Treasure Valley in western ID. Mountains in northeastern ID we will call the
central ID National Forests.
41) Boise, ID (BOI) 12Z sounding 28 April 2003. Red line represents lifted parcel
based on forecast temperature / dewpoint. We use observed values to modify the
sounding for late afternoon. 1000 J/kg CAPE in modified sounding. Forecast LFC is
around 820 mb, height of freezing level around 760 mb. Recall that PIH CWA is at
a higher elevation, and therefore cloud base will be closer to the surface Notice that
there is a relatively deep positive area in the layer colder than -10 C indicating a good
potential for graupel production and CG lightning.
42) Surface observations overlaid with RUC 00 hr forecast MSLP for 14:00 – 18:00
UTC. Notice broad pressure trough extending from northern Utah to western Oregon.
Temperatures are on the cool side in Montana / western Wyoming where the
shortwave is still affecting the region and weak n to nerly wnds are flowing into fcst
area. To the southwest skies are clearing and temperatures are warming into the 50s.
Northeasterly winds over Montana, southerly winds are developing over Utah and
central Nevada.
43) Visible loop 15:30 - 18:45 UTC. Cumulus forming on ridgelines as shortwave
approaches. Clearing over Snake River Valley between the two shortwaves is
evident. Probably cool, northerly flow. Note cirrus across Utah, where the strongest
flow aloft (jet streak) is located (fcst area very close to exit region of jet)…also note
gravity wave moving toward area associated with the jet. The first short wave in
Idaho is rather slow to move out as it moves into the diffluent flow…while the second
shortwave to the s and w is moving much faster towards the fcst area.
44) SKIP: ETA - 500 heights, vorticity, isotachs shaded. 12Z - see vort max
associated with lead short wave and second vort max approaching. Analysis seems
very noisy compared with 6.7 m imagery. 18Z - see secondary short wave
approaching. 00Z - PVA over area of interest - good diffluence - area in left exit
quad of jet - 45 kt at 500mb. Also note southerly flow at 500mb. Note spurious,
vorticity minimum in SW ID. This is due to shear induced vorticity on the northwest
edge of the jet (note for those who have taken the VISIT Cyclogenesis session,
500mb vorticity can be misleading at times). :SKIP
45) ETA - MSLP, surface winds. 12Z through 0Z - with time, ETA pushes southerly
flow throughout eastern Idaho and convergence is in the far northern part of the CWA
along the southern end of the central ID National Forests.
46) Mesonet surface observations for 21 UTC over southeast Idaho, western
Wyoming, northern Utah. Northerly winds, cooler temperatures and higher dewpoints
throughout Snake River valley, “pushing” against mountains in S. Idaho - this is NOT
what the ETA forecasted.
47) ADAS (ARPS Data Analysis System) analysis of surface wind isotachs and barbs at
22:00 UTC. Convergence has set up in S. Idaho along the northern edge of the higher
terrain.. Recall that the Eta had predicted uniform southerly flow throughout this
region by this time.
48) VIS loop - 17:00 - 23:30 UTC. At 17:00 UTC the shortwave in Idaho is centered
near Boise and still forcing convection. The second shortwave triggers new storms
by 20:30 UTC in extreme southern Idaho - one storm travels east-northeast along the
convergence zone. Remember the mid-level jet max is situated over Utah. Which
storm is strongest and has the best chance of surviving? This storm (ERN) is moving
rapidly eastward, probably due to propagation on the terrain related boundary. At
22:15 UTC the inflow feeder bands on the southern edge of the largest storm. WRN
storms get cut off from moisture and move into relatively cool and stable air.
49) KPIH radar with lightning - 20:51 - 23:41 UTC overlaid on local terrain. Note
lightning data show some positives, but mostly negative. The modest flash count and
occasional positive production is probably due to somewhat limited moisture (DPs in
the 40s). Large hail and tornadoes are observed. The storm travels along the terrainanchored convergence line. Storm has large area of high reflectivity, approaching 70
dBz. Second storm to the northwest of the bigger storm also produces a tornado, but
weakens as it moves into cooler air and the lower terrain of the Snake River Valley.
The radar signature cycles between large round shape HP-like, and LP mesocylonic
inflow notch type storms. At 2300 UTC the southwest flank of the storm is in the
Arbon valley near the PIH airport.
50) Picture of tornado from the storm mentioned above near 2300 UTC.
51) Un-remapped GOES-12 VIS loop - 14:15 - 01:31 UTC. GOES-12 is in RSO
because of moderate risk in central plains. Better temporal continuity. Viewing
angle allows us to see inflow side of storm, beginning at 21:15. Note the flanking
line and some feeder bands around 21:30 UTC.
52) Conclusions
53) 10 Nov 2002 Case intro with Northern, IN (IWX), Cleveland, OH (CLE) and
Wilmington, OH (ILN) CWAs.
54) Water vapor loop with ETA 500mb heights - 05:45 - 14:32 UTC. Shortwave
moving through area of interest overnight, in association with large scale trough back
to the west. Shortwave doesn't look too impressive in model height field, but better in
satellite.
55) ETA MSLP, winds, CAPE (shaded) - 12Z through 0Z - Good CAPE (for November)
developing and moisture advection ahead of the front. The Eta did have precip
developing along the pre-frontal trough in Indiana / Ohio which may account for the
lower forecast CAPE values at 18:00 UTC.
56) VIS loop - 13:15 - 18:15 UTC. Remember our area of interest is currently Ohio and
Indiana. Storms going up in southern IL and IN. Note the boundary along the IL/IN
border at 17:32 UTC (ahead of the developing squall line) which has more vertically
developed cumulus towers. Between this boundary and the pre-frontal trough line the
streets appear more unstable. The streets are oriented SW-NE, while the streets ahead
of the line are more southerly indicating convergence toward the northernmost
portion of the line.
Your eye is drawn to a second organized cloud line crossing Tennessee. This
feature will be discussed later. There also appears to be a synoptic scale boundary
which stretches from West Virginia to Illinois moving northward. Initial thoughts
may be that this is the warm front, but surface observations indicate it is not. When
we increase the animation speed and utilize the rock feature, we can easily observe a
northwestward moving feature that begins at the southeastern tip of IL / southwestern
tip to IN. A thunderstorm forms as this feature intersects the pre-frontal trough where
enhanced Cu is already developing.
57) 12Z sounding from Wilmington (ILN) - moderate CAPE, a little low-level turning,
good shear. SPC had issued a large high / moderate risk area for this day. The
question that is important to the individual WFO is can we anticipate localized areas
where the near-storm environment points to more favorable regions. Now let’s look
at the satellite imagery to key on features identified earlier.
58) VIS loop with 5-min CG lightning overlay - 17:40 - 20:45 UTC. Storms on northern
end of the line more intense as supported by the higher flash count. Note the
lightning data also hint at a northeastward moving storm ahead of the line. There are
multiple curved convergence lines ahead of the convection...northern end obscured by
the anvils. Convection seems to be developing along the northern end of these
convergence lines. At around 20:30 UTC, F4 Tor in nwrn OH…long track violent
tornado.
59) Ohio radar composite with surface obs - 17:36 - 20:42 UTC - Around 20:00, storms
close enough to the radar so we can see the curved convergence line, which is clearly
connected to the strongest storm in northwest Ohio. The surface data in southwest
OH do not reflect the passage of this feature, indicating it may be elevated. You
should always be suspicious of bands of organized convection that persist for a
long period of time and get deeper with heating, especially when they are
connected with a squall line.
60) Nashville (OHX) and Morristown (MRX) Tennessee CWAs. The purpose of this
short segment is to learn to recognize and track a mesoscale feature that turns out to
be important to convection later in the day, and was not handled well by the model.
In other words, to locate model unidentified meso/beta scale disturbances with
satellite imagery.
61) 12Z sounding at OHX – verifies high CAPE and reasonable shear but again the
question is can we use the imagery to identify localized regions that are more
favorable.
62) VIS loop - 13:15 - 22:10 UTC – This loop shows a mid-level disturbance (possible
mesovortex) moving into western Tennessee. This mesoscale feature is ahead of the
cold front, and situated between two shortwaves that were identified on the 12:00
UTC model runs. It clearly plays roll in initiation over central Tennessee. We won’t
be showing the model output, but this feature was hinted at on the RUC 700mb
analysis. Note that the small wave also seems to play a roll in destabilization of the
airmass in central TN. Cumulus is suppressed for most of the morning, then streets
and growing cumulus become evident around 1900 UTC. Unstable looking cumulus
move northward and intersect the storm in central TN around 21:15 UTC.
63) OHX/MRX storm relative radar loop with lightning overlaid - 20:28 - 01:00 UTC.
The lighning data show a high percent of positive CGs through 22:05 UTC. The
storm develops a higher flash count at approximately this time. The increase in flash
count, while maintaining a high positve count in the updraft, preceeds the first
tornado touchdown at 22:18 UTC. Near the end of the loop (see for example 00:43
UTC) feeder bands can be seen on the southeast flank of the storm as the storm gets
closer to the radar at KMRX. Here are the report during this period.
22:18 – tornado (PSD)
22:30 – wind damage (PSD)
22:35 – hail (PSD)
23:15 – tornado (High flash count)
23:22 – tornado
23:30 – tornado (2 dead)
23:34 – back to NSD
64) Jackson, MS (JAN), Birmingham, AL (BMX ), and Atlanta, GA (FFC) CWAs. The
purpose of this section is to review shortwave identification and interactions between
the near storm environment and the storms (first part), and to show how terrain can
play a role even over the southeast U.S. (second part).
65) Water vapor loop, 19:02 – 23:15 UTC, 10 Nov 2002 over southeastern U.S. Note
secondary shortwave that moves from north-central OK into IL. Enhanced WV
ahead of this shortwave moves into northwestern MS by 20:30 UTC. Note the
enhancement on northern edge of synoptic jet as the wave moves from northeast OK
into MO. Also note the small dry slot at the back edge of this wave that moves from
north-central TX into southeastern AR by the end of the loop. Find shortwaves in the
distorted flow…nearly baroclinic leaf/cusp looking (cyclogenesis referral).
66) 3.9 m shortwave IR loop - 20:45 - 0015 UTC. Near, and especially after, sunset we
advise looking at the 3.9 m imagery. 3.9 m is better than 10.7 m, because the
information about cloud fields is enhanced by the fact that water clouds and ice
clouds can be discriminated. Using this enhancement, the low level clouds show up
as dull gray, the water/ice mix clouds as dark gray, and ice clouds as the darker colors
or noisy-looking bright gray green. Note mesoscale deck of clouds that forms in
southeastern AR and northwestern MS in association with the shortwave. This is a
field of altocumulus castellanus (10.7 m and 00Z JAN sounding). As the leading
edge reaches central MS, a line of storms erupts and the accas dissipates. Why?
Note organized bands of low-level cu developing in southwest MS. There are also
many SW/NE oriented lines in AL and GA which may play a role later. NOTE:
There will be slightly better resolution on Ch-2 at night, due to less diffraction.
67) Series of soundings from JAN taken at; Toggle 1) 12:00 UTC, 10 Nov 2002, Toggle
2) 18:00 UTC, 10 Nov 2002, and Toggle 3) 00:00 UTC, 11 Nov 2002. Series
illustrates low-level inversion (cap) lifting and low-level winds backing as shortwave
approaches.
68) Toggle 1) Topography with surface observations overlaid. Note the dewpoint
gradient across central AL corresponds to a region where terrain is increasing.
Higher terrain over northeast GA is the southern end of the Blue Ridge Mountains.
Slightly lower dewpoints in northern AL are a common feature under most
circumstances according to the BMX office. By fading between the GOES 3.9 m
IR imagry and the terrain, we can see the relationship between the low-level
organized lines of Cu and many of the terrain features.
69) Columbus, AFB, MS (KGWX) 0.5 degree tilt reflectivity - 23:22 - 02:00 UTC. This
loop shows the early outcome, just after the storms develop,and before the activity
begins to line out.
1) Fist storm to cross into AL develops a hook, and produces tornado around
00:45.
2) The next storm coming into view in E. Mississippi, produces an F3 tornado.
70) SKIP: Storm relative loop - velocity with lightning - 23:37 - 01:56 UTC. Loop is
artificially produced, but allows us to track the first storm and display a couple of
important characteristics. There is a low total CG count in the beginning, but with an
elevated percent positive. The count increases significantly around 23:50. Shear
signature first appears around 00:13, and a good mesocyclone around 00:31. Large
and damaging tornado is reported around 00:45 – 01:01. It is possible that the
tornado report time lagged from the first actual touchdown, since it is night. :SKIP
71) GOES 10.7 m IR loop - 01:02 – 10:31 UTC 11 November 2002. Loop shows MS
activity lining out as it moves east. As the squall line develops, we can still observe
individual OST’s for quite some time. Also note the convection developing ahead of
the line indicating increasingly unstable air mass. Mergers between the pre-squall
line convection and the main line of storms occurs in north GA. If you vary the
animation speed you can identify the mountains in northeast GA and enhanced
brightness near the area of increasing terrain (i.e., careful observation shows a slightly
cooler brightness temperature region oriented WSW to ENE along the south side of
the higher terrain [e.g., 01:31 UTC]). A close look at the low-level flow and terrain
shows this phenomenon to be moist air pooling at the terrain barrier.
72) Terrain over northern Georgia overlaid with 01:00 UTC 11 November 2002 surface
observations. Strong southerly flow is damning along the southern rim of the
Appalachians (moist air damning). Remember the 10.7 m IR showed the moist
boundary layer may be deeper in this region (refer to previous slide). Note the strong
southerly winds further south and lighter winds beneath the dammed air mass. The
counties affected are Cherokee, Pickens, Dawson, Lumpkin and northern Hall.
73) Atlanta, GA (KFFC) 0.5 degree tilt reflectivity 06:48 – 08:11 UTC 11 November
2002 with lightning (toggle on) and 1200 foot contour (toggle on) overlays. Notice a
number of cores embedded within the eastward moving squall line. Convection is
moving ahead of the squall line toward the northeast. As the line reaches the deeper
moist boundary layer one cell intensifies. Deeper boundary layer along srn edge of
high terrain…taps and keeps tapped into the moisture/energy source…intensifying
storms. This what occurred.
06:48 – tornado in N. Cherokee county (just south of Pickens county border) - F2
06:54 – tornado in Pickens county (F2)
07:25 – tornado in Dawson county (F2) 13 total injuries
74) KFFC 0.5 degree tilt velocity 06:48 – 08:11 UTC 11 November 2002 with lightning
toggle. Highlighted counties are from west to east Pickens, Dawson and Lumpkin.
After a series of tornados in these counties, the storm goes on to produce severe wind
damage further northeast in Lumpkin county. Lightning is showing signs of pulsing
behavior and high positive flash count.
75) November 10 conclusions
76) Case study: 20 March 1998. Roanoke, VA (RNK) and Raleigh-Durham, NC (RAH)
CWAs.
77) Water vapor loop - 08:30 - 16:45 UTC. There is a cutoff low over W. Tennessee,
negatively-tilted trough over the southeastern U.S. Good diffluence to the east. Dry
slot wrapping around system from the south, approaching NC. If we rock the loop at
high speed we can easily see the shortwave rotating through southern MS/AL, into
GA then SC.
78) ETA - 500mb heights/winds - 12Z analysis and 18Z forecast - diffluent, but tight
gradient so winds are strong (60 knots). Heights are dropping because system is
slowly moving east.
79) Greensboro, NC (GSO) sounding 12Z - deep moisture - warm layer near 500mb will
be cooling with time - inversion weak, so storms may go early - wind profile shows
good deep layer shear with directional turning in the low-levels.
80) ETA - CAPE, SR-helicity, and 6km mean wind, convective precip - 12Z analysis, 6
and 12 hr forecasts: CAPE axis extends from central SC to central NC, 18Z: CAPE
increasing across SC, but relatively low in Central NC, because Eta develops
convective precipitation. SRH over 200.
81) Surface obs - heating more than ETA forecasted - surface winds stronger than RUC
forecasted. Where is the warm sector? Answer: Generally over central NC with
stable air to the east and northeast. Storm motion will keep storms moving within the
warm sector rather than moving east into the stable air. Note the good dry punch
intruding from the west, pushing towards the area northwest of Greensboro.
82) Visible loop – 13:15 - 18:15 UTC – The visible imagery makes analysis of the warm
sector simpler and more precise. Compare the 18:00 UTC visible image with the
corresponding surface obs. Instability in the warm sector is implied by clearing skies
and developing cumulus through SC into central NC. Note that the precip the ETA
had forecast is not developing, though there is a line of storms forming in western
Carolinas. Looping quickly you can see stratiform stable looking far to the east, so a
narrow tongue of high CAPE (at least higher than the Eta forecast).
83) Visible loop - 17:20 - 23:15 UTC - point out clearing line - convection begins looking
like it'll be linear, but with time, individual storms become evident in the line, so it's
not really linear. Note the apparent bulge in the line of storms in the vicinity of the
dry push. Also notice the convective line(s) extending from extreme eastern SC into
north central NC. Tornado touchdown at 20:13 (F1) and at 20:25 (F3) on same storm
on the storm just north of a "break" in the line...2 primary storms evident at this time
(northern storm produced tornado). at 19:55, look for feeder clouds on the SW flank
of that storm. Large outflow developing also. at 20:15, signs of an RFD when
multiple lines of cumulus have developed on the western flank of the storm.
84) Blacksburg, VA (KFCX) radar 0.5 reflectivity with surface obs - 18:37 – 20:32 UTC
– F3 tornado is observed at 20:25 near the NC/VA border. Reflectivity alone would
suggest that tornadic storm and the storm to its south are approximately the same
intensity – in fact, the southern core has slightly higher reflectivity. We will look at
velocity soon, but first let’s see if there are any clues in the infrared imagery.
85) IR channel 4 loop - 18:55 – 20:35 UTC. The infrared loop shows that the coldest
tops continue to be found in the area of the dry push in north central NC and southern
VA. The northernmost storm is clearly the storm with the coldest top, indicating that
it has the tallest updraft. This storm also exhibits an enhanced-V signature.
86) KFCX radar 0.5 degree base velocity at 20:27 UTC. The northernmost storm is
clearly the storm with the better velocity couplet. This is the storm that produced the
F3 tornado around 20:25 UTC.
The theme here is that RSO satellite data can be used to get additional clues about
storms. Imagine a scenerio where the storm is far from the radar or a radar is down,
by having the RSO schedule in effect you can make use of the data to get information
about the storm as well as the environment around the storm. This can only be
accomplished in the RSO schedule as the 15-minute schedule does not appear on your
AWIPS in sufficient time to make use of it in this way.
87) March 20 conclusions
88) Quiz intro page
89) 1900 UTC visible image with surface overlay on 4 May 2003 – Beginning of a
long WEEK of Severe Wx. Question: ID air masses. Describe synoptic features for
eastern Kansas and northeast Oklahoma (Note: make sure that students mention two
northeast/southwest oriented convergence lines - "double dryline", with 40's dew
points west of the westernmost line and a dew point gradient across both lines).
Towering cumulus have already begun forming on the northern end of the
easternmost line.
90) Visible loop - 18:45 - 23:02 UTC - Start the loop. At 20:15, notice break between the
two lines in northeast Oklahoma, along with KS/OK border, and the storm forming
on the southern end of the eastern line, (storm 1). Notice that this storm is moving
into air which has cloud streets in SE Kansas and SW Missouri.
Definitions: (NS) Northern storm (the storm that dies). (SS) Southern storm will be
the one that’s intensified. (LM) Left mover that intersects both storms.
Exercise: Point out splits, mergers, and anomalous motion.
One storm splits at 21:15 UTC in southeast KS.
SS forms along the westernmost, or the true, dry line in northeast Oklahoma, and
by 21:55, is moving east. It has feeder bands and an overshooting top.
left-moving storm in OK (LM) forms from a split and can be seen approaching
the area of interest beginning around 21:45 UTC. This storm is moving northeast at
70 knots.
Question – How does the shape of the anvil on LM differ from the others?
Why is that? (Vector difference between the anvil lvl wnds and the storm
mostion. )
LM merges with SS at around 22:45 UTC, overrunning huge feeder bands on the
southeast flank of SS
Note that SS has the best access to the air containing the cloud streets
91) SGF base reflectivity with 5-min CG lightning - 20:32 - 00:48 UTC Question: what radar and lightning characteristics suggest these storms are severe?
NSD storms, in general, are producing greater 15% positive CG, with overall high flash
rate and a pulsing flash rate. Note the storm split in SE Kansas near beginning of
loop. The right mover is PSD, while the left mover is NSD.
LM has little or no lightning as it approaches SS. SS has about 70-80 strikes per 5
minutes (15% positive). As interaction begins, total strike count and % positive both
increase in SS. As LM moves northward out of SS the lightning count in SS goes
back to what it was before the interaction but the % positive is greater. Before LM
passes in front of NS strike counts are roughly 50 strikes in 5 minutes with a high %
positive. As LM crosses in front of NS, NS weakens, dies, and lightning decrease
precedes this by about 5 minutes. Just after 23Z – tornado in nern OK – ottowa
county.
Definitions: (NS) Northern storm (the storm that dies). (SS) Southern storm will be the
one that’s intensified. (LM) Left mover that intersects both storms.
Note that the fast moving LM interacts with SS in extreme SW Missouri around 23:00
UTC (noted on satellite loop above). It intersects this storm near SS updraft and
FFD. After this interaction, southernmost storm appears to weaken for a time as it
crosses into southeastern Missouri, but then reorganizes and eventually develops a
long-track, devastating tornado (Pierce City). LM continues north and crosses the
path of NS, cutting off its infow. NS dies. On that day – 4th of May – there were
38 tors in Missouri alone – 4 of which where F4s!!! … and between the 4th and
the 6th of May(3 day total) … 108 tors…just in Missouri!!
Other: 21:56 UTC 3” hail with SS. 22:55 UTC tornado with NS and SS. 23:05 UTC
4 ¼” hail with SS. 23:16 UTC tornado with NS. 23:24 UTC 1 ¾” hail with SS.
00:04 UTC large tornado with SS. 00:09 UTC tornado with SS. 00:23 UTC tornado
with SS. 00:26 UTC tornado and 2 ½” hail with SS. 00:45 UTC tornado with SS.
92) Two days later: 6 May 2003 - ETA - dew point, MSLP, winds - by 12Z, sfc low
over C. Oklahoma, moisture advecting north, cold air coming south thru W. Kansas.
18Z - moisture indeed advecting north. We need to use satellite and surface obs to
track the progress of the moisture return.
93) VIS with Sfc data (toggle)- 11:15 - 16:55
Question: There is a shortwave trough moving thru KS - what are some signs of this
shortwave in the VIS?
Answer: Cloud mass in central KS with convection on leading edge, jet cirrus streaming
in south of the wave. Note: there is cooler, dryer air at the sfc, convection forming
over this (elevated convection).
Question: Step thru and look for northward moisture advection in satellite and/or surface
air.
Answer: note the dew point go up with the passage of the northwestward moving line of
low clouds...we can now track this more moist air
Question: What effect does the arrival of this moisture have on the elevated convection?
Answer: storm develops inflow bands, gets larger, and develops a large flanking line
(becomes surface based and stronger). Tornado reported at 15:48 as moisture
intersects storm.
94) VIS loop - 13:31 - 00:15 UTC
Question: Storm on southern end of the line forms at 1945 - when does it look severe?
Answer: flanking lines and feeder bands at 2115 - that's when it tornadoes in Allen
County Kansas...this continues into Bourbon County and then into Missouri. Another
storm to the northeast at this same time also has feeder bands. Also notice the
southward moving outflow boundary which aids in the initiation of storms along the
southwest/northeast oriented line. Another big tor day in Missouri with 34 tors on
the 6th.
95) SGF radar - base reflectivity - 20:59 - 00:19 UTC.
Question: find outflow boundary and discuss what effect it has when intersecting other
storms.
Answer: more stable air behind the boundary causes the tornadic storm moving across the
KS/MO border at 22:00 UTC to weaken. It intersects eastward moving storm in SE
Missouri causing it to develop a nice hook echo and it tornadoes.
96) Visible loop for June 24, 2003 case. Identify the various air masses. Enhanced
cumulus line moving east, storms later form on this. Points to notice: 1) cloud streets
throughout most of eastern NE (parallel to low-level flow), 2) old MCS boundary
from n-central NE to se-SD to southern MN with stable wave clouds to the north of
the boundary (oriented perpendicular to flow at top of boundary layer, 3) stratoform
clouds in western NE with warm frontal boundary just east of stratiform field (note:
there is a low in ne-CO).
The MCS boundary moves north with time (fast/rock), possibly merging with warm
front. Organized n-s line of enhanced Cu in eastern NE merges with warm
front/MCS boundary and a storm forms in se-SD. Storm becomes tornadic almost
immediately (first report at 21:30 UTC). Second storm forms on Cu line in eastern
NE and splits at around 22:30 UTC. Several tornado reports near 23:00 UTC near
Missouri River. Meanwhile, the original MCS boundary storm produced intense
outflow boundary as it dissipates and a new storm form. A large tornado touches
down in Sanborn county SD at 23:19 UTC with this storm. Note that a thunderstorm
forms at around 20:30 UTC near OMA NE. That this storm is blocking the low-level
flow can be seen with the dissipation of the clouds streets north of the storm
beginning almost immediately. The effect is particularly noticeable at 22:45 UTC.
Also, notice that the storms in nc-IA that move into the clear area, dissipate.